Once a patient has undergone an organ transplant, it is crucial that the patient's immune system be partially disabled so that it will not reject the allograft. The immune system must be disabled enough to prevent the body from rejecting the graft, i.e., mounting an adverse immune response, but not so much that the immune system will not be able to defend against viruses, bacteria or other infection. This balance must continue throughout the remainder of the patient's life to assure that neither rejection nor undue susceptibility to infection will occur.
One immunosuppressant drug utilized in the prior art for partial disablement of the body's immune system is Cyclosporin A ("CyA"). CyA has been shown to be effective in targeting and partially disabling T cells, which are responsible for fighting off alien tissue, thus allowing for more successful transplants. CyA, however, tends to inflict damage on various organs due to its high potency. In order to ameliorate the harsh effects of CyA, the dosage of CyA administered to a patient in the prior art is minimized, and each dose is supplemented with dosages of other drugs such as steroids.
Recently, it became known that a new drug, FK 506, which is 100 times more potent than CyA in inhibiting lymphocyte proliferation in mixed lymphocyte cultures, can be used as an immunosuppressant in place of CyA to prevent rejection of organ transplants. FK 506, a macrolide antibiotic produced by the fungus Streptyces tsukubaenis, has clinically been found to have a potent and targeted effect on T cells without the severe range of side effects exhibited by the similar use of the drug CyA. In addition, the use of FK 506 rather than CyA has been shown to lessen and even eliminate the need for the continued use of steroids. This, in turn, results in reduced instances of hypertension in transplant patients. R. Venkataramanan et al., Pharmacokinetics of FK 506 in Transplant Patients, Transplantation Proceedings, p. 2736 (December 1991).
FK 506 has also been found to be successful with patients afflicted with autoimmune diseases. As in the case of transplant patients, FK 506 inhibits the patient's T cells, prohibiting them from attacking the patient's own body, i.e., mounting an adverse immune response. Autoimmune diseases include, inter alia, multiple sclerosis, diabetes, psoriasis and rheumatoid arthritis.
Presently, other than a trial-and-error technique--inappropriate for FK 506 treatment (for the reason discussed below)--there are no known techniques for treating patients generally with FK 506. There is no standardized set of patient criteria which one can examine in order to determine what course of action to take with respect to treating a patient with FK 506, and there is no set of standardized characterizations for any such criteria. Moreover, there is a large inter-individual variability in the pharmacokinetics of FK 506. R. Venkataramanan et al., Pharmacokinetics of FK 506 in Transplant Patients, Transplantation Proceedings, pg 2736 (December 1991). See also V. S. Warty et al., Practical Aspects of FK 506 Analysis (Pittsburgh Experience), Transplantation Proceedings, pg 2730 (December 1991); G. J. V. Nossal, Summary on the First International FK 506 Congress: Perspective and Prospects, Transplantation Proceedings, pg 3373 (December 1991). Thus, what may be an appropriate course of action with one patient may not be the appropriate course of action with another patient. Therefore, prior to the present invention, a physician needed to do a substantial amount of experimentation with FK 506 on a patent before the physician was in a position to know what course of action to take with respect to treating that patient with FK 506.
The study and development of artificial intelligence ("AI") has recently provided a means for simulating the human decisional processes and has developed to a point where it can now be used in certain situations to solve problems and yield much the same results as human beings would. Thus, AI can help reduce the learning curve associated with acquiring knowledge in a particular area, including drug treatment.
An expert system is an application of AI designed to solve problems through the manipulation of data. Expert systems typically include a knowledge base of data and rules relating that data to additional data which is fed into the expert system for purposes of determining the solution to a problem.
Expert systems have recently found use in a variety of applications, such as agriculture, chemistry, computer design, construction, engineering, finance, management, health care and manufacturing. Moreover, prior art expert systems have been designed to address a relatively wide range of health care concerns. See Dormond et al., U. S. Pat. No. 4,839,822.
Beckers, U. S. Pat. No. 5,019,974, discloses a computerized system for diabetes management which gathers, processes and analyzes data to provide an outpatient with a personalized tailored treatment or medication program. Dormond et al., U. S. Pat. No. 4,839,822, discloses an expert system for use in treating various types of trauma.
Nonetheless, there are presently no expert systems known to the applicant that have been developed for use in treating patients to prevent an adverse immune response, whether or not that treatment involves FK 506.
Due to its potency, an overdose of FK 506 can fatally cripple the immune system's ability to defend against bacteria and viruses. An under-dose, on the other hand, can result in an adverse immune response in a patient who has received a transplanted organ or is affected with an autoimmune disease. It is, therefore, critical that a patient who requires FK 506 be given the proper dosage so as to balance these harms and that therapeutic monitoring occur. See J. McMichael, Evaluation of A Novel "Intelligent" Dosing System For Optimizing FK-506 Therapy, Transplantation Proceedings, p. 2780 (December 1991).
Determining the proper FK 506 dosage to be administered to a patient generally is complicated by the inter-individual variability in the pharmacokinetics of FK 506. Moreover, the amount a given patient needs changes over time depending on how close the patient is to being stable. A patient is considered stable when he or she no longer exhibits the effects of receiving too much FK 506, known as toxicity (in particular, nephrotoxicity), and displays no rejection or autoimmune disease flare ups.
None of the methods presently available for dosing patients is capable of handling these variables. Thus, none of them can be used to determine the proper FK 506 dosage to be administered to a patient at a given time.
Linear equations based on population parameters such as age, weight or sex assume that people respond to a particular drug in accordance with their population characteristics. That assumption, and those types of equations, fail to account for variability among people within the same population parameters.
Non-linear least squares modeling methods cannot be used either. Those methods involve use of a vast amount of data concerning the general population as a whole to arrive at a solution. The larger the quantity of data, the better the fit. Not only do those types of methods involve the use of unnecessary criteria, but moreover, like linear equations, they fail to account for variability among people within the same population parameters.
Bayesian analysis is also inappropriate for determining proper FK 506 dosages. Unlike linear equations and non-linear least squares methods, Bayesian analysis does employ specific data about the medical status of a particular patient. However, the analysis also requires results from previous dosing experiences with the general population. As already mentioned, relying on the general public to determine the appropriate dosage of FK 506 for a particular patient is inappropriate because of the inter-individual variability in the pharmacokinetics of FK 506. Moreover, Bayesian analysis's ability to determine a proper dosage becomes reliable only after treating a patient with a particular drug for a significant period of time. Thus, it cannot be used to accurately predict proper dosages during the early dosing periods of a drug, especially FK 506.
Pharmacokinetics compartment modeling is not a viable way of dosing a patient with FK 506 either. While R. Venkataramanan (in his paper entitled Pharmacokinetics of FK 506 in Transplant Patients published in the Transplantation Proceedings in December of 1991 at page 2736) has suggested that a two compartment model could be employed, a subsequent experiment, which utilized such a model for dosing FK 506, shows that compartment modeling will not work.
The problem with pharmacokinetics compartment modeling is that the model comes from a standardized equation which is modelled upon the general population's interaction with a particular drug. It, therefore, fails, once again, to take into account inter-individual variability. (An example of a one-compartment pharmacokinetics model for dosing aminoglycosides can be found in Abbott Diagnostics' Aminoglycosides dosing computer program and instruction manual published in 1988. An example of an expert system which uses pharmacokinetics compartment modeling is Zarawitz et al., U. S. Pat. No. 4,888,014.)
The final method presently available for determining FK 506 dosages is the trial-and-error method. That is, a patient could be given incremental dosages of FK 506, and the patient's reaction thereto could be observed and used to determine the frequency and quantity of subsequent dosages. The risks of a trial-and-error methodology of prescribing FK 506 are, however, evident. An overdose would be extremely difficult to avoid and, as mentioned before, could cripple the immune system's ability to defend against bacteria and viruses. An under-dose could result in rejection or a relapse of an autoimmune disorder.